“Push synchronization to the edge of the reactive graph”
TL;DR: Is this a thing?
In the Clojure community there was a saying:
“Push your impure functions to the edge of the system”
Working with
cache
and createAsync
I'm moving towards
“Push synchronization to the edge of the reactive graph”
- have a synchronization area where the up-to-date data exists.
- have external events update the synchronization data and then force revalidate the cache
points depending on that data.
- share the cache
points via context (rather than a reactive value).
- In the component mount createAsync
or createAsyncStore
(with the appropriate reconcile
options) on the cache
point to get the reactive behaviour.
Is this a thing or am I seeing things?
One benefit seems to be that stores get a lot easier to deal with as you just update the synchronization data wholesale and leave the grunt work to reconcile
.
Also you can have multiple cache
points depending on different parts of the same data.0 Replies